[BUG] linux-cachyos 7.2.0/7.2.2: amdgpu DC REG_WAIT timeout (optc314_disable_crtc) causes black screen / greeter crash on Radeon 780M (HawkPoint); LTS and 7.1.8 work

### Pre-flight checklist

- [x] I have searched existing issues and this is not a duplicate.
- [x] I have read the [Contributing Guidelines](https://github.com/CachyOS/linux-cachyos/blob/master/CONTRIBUTING.md).
- [x] I have verified the issue is reproducible with the latest available CachyOS kernel.
- [ ] I have tried to reproduce the issue on Arch Linux's `linux` kernel. *(not tested — no vanilla Arch kernel installed on this system)*

### Upstream / vanilla kernel check

I have not tested with a vanilla/upstream kernel

### Kernel variant

linux-cachyos (EEVDF, Clang)

### System information (cachyos-bugreport.sh)

`cachyos-bugreport.sh` requires root and wasn't available to run in this session — manual system information provided below instead.

### Manual system information (if cachyos-bugreport.sh is unavailable)

```text
Device: Micro Computer (HK) Tech Limited "EliteMini" (Minisforum UM870)
CPU: AMD Ryzen 7 8745H w/ Radeon 780M Graphics (HawkPoint)
GPU: integrated AMD Radeon 780M, amdgpu driver (PCI 0000:c4:00.0, IP DISCOVERY 0x1002:0x1900 0x1F4C:0xB016 0xB3)
RAM: 13 GiB
Arch: x86_64 (x86-64-v4 package build)
DE/session: KDE Plasma, Wayland (plasma-login greeter/SDDM replacement)
Filesystem: Btrfs
sched_ext: disabled

Working: linux-cachyos-lts 6.18.42-1, linux-cachyos 7.1.8-1
Failing: linux-cachyos 7.2.0-1, linux-cachyos 7.2.2-1
```

### Bug description

After upgrading `linux-cachyos` from 7.1.8-1 to 7.2.0-1 (2026-08-24), the system stops reaching a usable graphical login. The kernel boots and all other services (network, docker, etc.) start normally, but the amdgpu display core fails to bring up any CRTC output, so the Plasma login greeter either shows only a placeholder screen with no real output, or crashes outright.

Confirmed still present after upgrading further to 7.2.2-1 (2026-08-29) — same underlying `amdgpu` REG_WAIT/blank-crtc errors, just surfacing slightly differently (greeter process crash with signal 1 instead of an indefinite placeholder screen).

Both `linux-cachyos-lts` (6.18.42-1) and the previous `linux-cachyos` 7.1.8-1 boot and display normally on the same hardware — this is a regression introduced somewhere between 7.1.8-1 and 7.2.0-1, and not yet fixed as of 7.2.2-1.

### Steps to reproduce

1. On a system with an AMD HawkPoint APU (Ryzen 7 8745H / Radeon 780M, amdgpu), have `linux-cachyos` at 7.1.8-1 or earlier (working) and `linux-cachyos-lts` installed side by side.
2. Upgrade `linux-cachyos` to 7.2.0-1 or later (reproduced on 7.2.0-1 and 7.2.2-1).
3. Reboot and select the `linux-cachyos` (non-LTS) boot entry.

### Expected behavior

System reaches the Plasma login greeter / SDDM with a normal display, same as on `linux-cachyos-lts` and the previous `linux-cachyos` 7.1.8-1.

### Actual behavior

- On 7.2.0-1: kernel and amdgpu load and detect the GPU/VRAM correctly, but DC (Display Core) fails to bring up any CRTC output. The greeter repeatedly logs "There are no outputs - creating placeholder screen" and the screen stays effectively black. System is otherwise fully booted (network, docker, etc. all running).
- On 7.2.2-1: same underlying amdgpu REG_WAIT timeouts, but this time the plasma-login greeter helper process crashes outright (signal 1) instead of showing a placeholder screen.
- Each affected boot lasts roughly 15–40 seconds before being rebooted back to LTS.

### Logs / stack traces

linux-cachyos 7.2.0-1 (2026-08-24 22:49) — kernel log excerpt

```text
amdgpu 0000:c4:00.0: initializing kernel modesetting (IP DISCOVERY 0x1002:0x1900 0x1F4C:0xB016 0xB3).
amdgpu 0000:c4:00.0: [drm] Display Core v3.2.384 initialized on DCN 3.1.4
amdgpu 0000:c4:00.0: [drm] Registered 4 planes with drm panic
[drm] Initialized amdgpu 3.64.0 for 0000:c4:00.0 on minor 1
amdgpu 0000:c4:00.0: [drm] Mode Validation Warning: Viewport size failed validation.
(repeated x6)
fbcon: amdgpudrmfb (fb0) is primary device
amdgpu 0000:c4:00.0: [drm] fb0: amdgpudrmfb frame buffer device
...
[drm:dcn20_wait_for_blank_complete [amdgpu]] *ERROR* DC: failed to blank crtc!
amdgpu 0000:c4:00.0: [drm] REG_WAIT timeout 1us * 10 tries - optc3_lock line:130
amdgpu 0000:c4:00.0: [drm] REG_WAIT timeout 1us * 100000 tries - mpc2_assert_idle_mpcc line:481
amdgpu 0000:c4:00.0: [drm] REG_WAIT timeout 1us * 100000 tries - optc314_disable_crtc line:148
amdgpu 0000:c4:00.0: [drm] Mode Validation Warning: Viewport size failed validation.
```

Userspace side (`plasma-login-greeter`), repeated every ~10s until reboot:
```text
plasma-login-greeter[928]: There are no outputs - creating placeholder screen
plasma-login-wallpaper[929]: There are no outputs - creating placeholder screen
plasma-login-greeter[928]: Creating a layer shell for placeholder screen. This will be positioned incorrectly
plasma-login-wallpaper[929]: Creating a layer shell for placeholder screen. This will be positioned incorrectly
```

linux-cachyos 7.2.2-1 (2026-08-29 10:28) — kernel log excerpt

```text
amdgpu 0000:c4:00.0: amdgpu: [drm] Display Core v3.2.351 initialized on DCN 3.1.4
...
amdgpu 0000:c4:00.0: [drm] REG_WAIT timeout 1us * 100000 tries - optc314_disable_crtc line:145
...
[drm:dcn20_wait_for_blank_complete [amdgpu]] *ERROR* DC: failed to blank crtc!
amdgpu 0000:c4:00.0: [drm] REG_WAIT timeout 1us * 10 tries - optc3_lock line:130
amdgpu 0000:c4:00.0: [drm] REG_WAIT timeout 1us * 100000 tries - mpc2_assert_idle_mpcc line:481
amdgpu 0000:c4:00.0: [drm] REG_WAIT timeout 1us * 100000 tries - optc314_disable_crtc line:148
amdgpu 0000:c4:00.0: [drm] Mode Validation Warning: Viewport size failed validation.
```

Userspace side, greeter crashes outright this time:
```text
plasmalogin[816]: Error from greeter session: "Process crashed"
plasmalogin[816]: Auth: plasmalogin-helper (--socket ... --start /usr/bin/startplasma-login-wayland --user plasmalogin --greeter) crashed (signal 1)
plasmalogin[816]: Error from greeter session: "Process crashed"
plasmalogin[816]: Greeter stopped. PLASMALOGIN::Auth::HELPER_OTHER_ERROR
```

### Additional system information

Possibly related (different kernel version / hardware, but same amdgpu DC/pageflip failure family, both closed): #959, #970. This report is for the same class of bug reproducing fresh on 7.2.0-1 and 7.2.2-1 on a HawkPoint (Radeon 780M) single-iGPU system, so filing separately in case root cause differs.

Happy to test any diagnostic kernel params, dmesg with `drm.debug`, or pull `cachyos-bugreport.sh` output if root access is available at time of testing (currently running LTS as my daily driver since non-LTS kernel is unusable).